Abstract: Through the use of information technology, this study combined computer vision and artificial intelligence （AI） to establish a system for the automatic recognition and control of driving behavior. By leveraging the autonomous learning and deep learning capabilities of AI, the system assisted video surveillance and visual analysis to automatically identify risk behaviors of drivers during operation, such as fatigue, distraction, and failure to wear seat belts. It provided proactive warnings and alerts through methods like voice notifications and seat vibrations, while also transmitting warning information to the dispatch room to remind personnel to analyze and intervene, enabling them to take effective measures in advance to prevent accidents. The system conducted real-time monitoring of drivers’ behaviors during driving, automatically assessed their mental state, and accurately identified common unsafe driving behaviors, providing proactive warnings and saving video records. Compared to manual daily management, this effectively resolves the issue of inconvenient inspections in the driver’s cabin during vehicle operation, enabling 24 h continuous operation and significantly improving work efficiency. The application of this system greatly standardizes the vehicle transportation operations, constrains unsafe behaviors of drivers, effectively prevents vehicle accidents, reduces the probability of traffic safety incidents in mining, and significantly enhances the inherent safety level of the mine.